A recent large clinical study demonstrated the association between intermediate CD14++CD16+monocytes and cardiovascular events. However, whether that monocyte subset contributes to the pathogenesis of atrial fibrillation (AF) has not been clarified. We compared the circulating monocyte subsets in AF patients and healthy people, and investigated the possible role of intermediate CD14++CD16+monocytes in the pathophysiology of AF. This case-control study included 44 consecutive AF patients without systemic diseases referred for catheter ablation at our hospital, and 40 healthy controls. Patients with systemic diseases, including structural heart disease, hepatic or renal dysfunction, collagen disease, malignancy, and inflammation were excluded. Monocyte subset analyses were performed (three distinct human monocyte subsets: classical CD14++CD16-, intermediate CD14++CD16+, and non-classical CD14+CD16++monocytes). We compared the monocyte subsets and evaluated the correlation with other clinical findings. A total of 60 participants (30 AF patients and 30 controls as an age-matched group) were included after excluding 14 AF patients due to inflammation. Atrial fibrillation patients had a higher proportion of circulating intermediate CD14++CD16+monocytes than the controls (17.0 ± 9.6 vs. 7.5 ± 4.1%, P < 0.001). A multivariable logistic regression analysis demonstrated that only the proportion of intermediate CD14++CD16+monocytes (odds ratio: 1.316; 95% confidence interval: 1.095-1.582, P = 0.003) was independently associated with the presence of AF. Intermediate CD14++CD16+monocytes were negatively correlated with the left atrial appendage flow during sinus rhythm (r= -0.679, P = 0.003) and positively with the brain natriuretic peptide (r = 0.439, P = 0.015). Intermediate CD14++CD16+monocytes might be closely related to the pathogenesis of AF and reflect functional remodelling of the left atrium.